A prognostic nomogram for colorectal cancer based on ubiquitin-specific protease 21 expression: a retrospective cohort study

Background Colorectal cancer (CRC) remains a leading cause of cancer-related mortality worldwide, and substantial prognostic heterogeneity exists among patients with similar TNM stages. Ubiquitin-specific protease 21 (USP21) has been implicated in tumor progression across multiple malignancies; however, its prognostic value in CRC has not been fully elucidated. This study aimed to evaluate USP21 protein expression in CRC tissues and develop a USP21-based nomogram for individualized prediction of postoperative overall survival (OS). Methods A total of 115 CRC patients who underwent radical resection at the Affiliated Kunshan Hospital of Jiangsu University between 2018 and 2019 were retrospectively included. USP21 protein expression in tumor and adjacent normal tissues was assessed by immunohistochemistry. Prognostic factors were screened using univariate Cox regression followed by least absolute shrinkage and selection operator (LASSO) Cox regression to construct a multivariable nomogram. Model performance was evaluated using the concordance index (C-index), time-dependent ROC curves, calibration plots, and decision curve analysis (DCA). Patients were stratified into tertile-based risk groups for Kaplan–Meier survival validation. Results USP21 expression was significantly higher in CRC tissues compared with adjacent normal tissues ( P < 0.001) and was associated with depth of invasion, lymph node metastasis, TNM stage, and lymphovascular invasion ( P < 0.05). Univariate Cox regression identified T, N, and M stages, LVI, PNI, and high USP21 expression as significant predictors of poor OS. LASSO-Cox regression retained five prognostic variables—USP21 expression, LVI, T stage, N stage, and M stage—for nomogram construction. Risk stratification based on nomogram-derived tertile cutoffs showed significant differences in OS among the low-, intermediate-, and high-risk groups (P < 0.001). The nomogram demonstrated strong predictive performance, with a C-index of 0.820 and time-dependent AUCs of 0.916 (3-year) and 0.854 (5-year). Calibration curves showed excellent agreement between predicted and observed survival, and DCA indicated substantial clinical net benefit. Conclusion The USP21-based nomogram integrating key clinicopathological variables provides accurate individualized survival prediction and serves as a promising adjunctive tool for refining postoperative prognostic assessment and guiding clinical decision-making.